Robust H^∞ optimal controller designs for feedback substitution schemes using LSDP and MOGA

摘要

We discuss a new methodology for designing optimal controllers tailored specifically for feedback substitution schemes. The loop shaping design procedure (LSDP) requires us to re-cast the problem using linear matrix inequalities to specify a range of objectives. We then use a genetic algorithm to perform a multi-objective optimization for the controller weights (MOGA). We contrast this methodology to that currently adopted to simultaneously minimize response time and noise variance in the feedback signal in simple SISO systems. We use robust control theory criteria to benchmark the performance of the designed controllers and compare them to those derived using analytical techniques.